Inhaled Virus-Mimicking Particles Elicit Dose-Dependent Immune Response against SARS-CoV‑2 with Sex-Bias ACE2 Activity in Mice Authors: Pei Zhang, Erik Niemelä, Panagiotis Filis, Pasi Sorvisto, Jani Virtanen, Hélder A. Santos Published: 2026 ABSTRACT There is an urgent need to develop antigen presentation platforms that can induce broadly systemic immunity against multiple severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) subvariants. Here, virus-mimicking particles (VMP) are functionalized with substantial receptor binding domain protrusions from Delta variant SARS-CoV-2 on particle surface, programmed by biotin–streptavidin interaction. The fabricated VMP reveal superior pulmonary biodistribution and retention for at least 1 day in C57BL/6 female mice after oropharyngeal aspiration exposure. VMP at higher dosage elicit a robust and broad systemic immune response against SARS-CoV-2 variants in K18 human angiotensin-converting enzyme 2 (ACE2) transgenic mice. Furthermore, ACE2 expression in males dramatically decreases after high-dose VMP inhalation, while that in females remains comparable to baseline, which might be the reason that females are less vulnerable to the adverse effects than males. Overall, VMP are a promising candidate for next-generation viral inhibitors and immune stimulating particles against current and future SARS-CoV-2 variants. The continuous emergence of highly mutated severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants highlights serious concerns about the antiviral efficacy of first-generation vaccines. The subsequently emerging variants of concern, including Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2), and Omicron (B.1.1.529, BA.2, BA.5, and XBB.1.5), are distinguished by their enhanced infectivity, pathogenicity, neutralization resistance, and immune escape. − The attenuated protection against the mutated variants provided by the first-generation vaccines can be mainly attributed to the deletions in the N3 and N5 loops and substitution constellations within the receptor binding domain (RBD). − Many of these mutations are shared by the Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), and Delta (B.1.617.2) variants. , Thus, there is an urgent need to develop next-generation vaccines that are safe and capable of inducing broadly protective immunity against multiple subvariants of SARS-CoV-2. The RBD is a portion of the S1 subunit at the apex of the Spike glycoprotein. The RBD possesses multiple conformation-dependent epitopes and can induce high-titer neutralizing antibodies. The highly conserved and vulnerable regions within the RBD preferentially focus the immune response toward potent cross-protective determinants. − For example, Liang et al. reported that the RBD trimer encoded by a lipid-nanoparticle-encapsulated nucleoside-unmodified mRNA vaccine was able to induce broad and protective immune responses against infection of wild-type SARS-CoV-2 and its major variants of concern in the mouse model. Thus, the RBD could serve as the precise target for the development of next-generation vaccines and immune stimulators that aim to induce broad-spectrum immune protection against SARS-CoV-2 subvariants. To be noted, nanoparticle trimeric RBD is capable of producing stronger humoral immune responses than mRNA trimeric RBD. Recently, we have developed virus-mimicking particles (VMP) through a surface biomimetic approach, efficiently blocking SARS-CoV-2 infection through competitive inhibition mediated by their interaction with the host ACE2 receptor. These VMP were decorated with an average of 231 or 444 RBD protrusions on each particle surface, replicating the wild-type SARS-CoV-2 or its B.1.617.2 (Delta) variant, respectively. Considering that natural infection was proven to be able to elicit a cross-reactive immune response, in this study, we further anticipated that the VMP decorated with RBD could serve as an inhaled broad-spectrum SARS-CoV-2 viral inhibitor and immune stimulator to replace traditional vaccines (Scheme ). To mimic the SARS-CoV-2 infection process, the VMP were administered to the lung through oropharyngeal aspiration. The VMP were supposed to indicate superior pulmonary biodistribution and retention because of its interaction with ACE2 and the resultant slower pulmonary clearance. We hypothesized that, once the inhaled VMP were internalized via interaction with ACE2, they could trigger robust systemic immune responses, stimulate the production of broad-spectrum neutralizing antibodies, and thereby elicit a broad and cross-reactive immune response against SARS-CoV-2 infection. Furthermore, we supposed that, similar to the SARS-CoV-2 infection process, VMP administration could induce sex-bias differences in ACE2 activity. Rationale Scheme Illustrating That Inhaled VMP Enable the Elicitation of a Robust and Broad Immune Response against SARS-CoV-2 Infection in Mice with Sex-Bias Differences in ACE2 Activity a(a) The VMP carrying a substantial number of RBD protrusions were fabricated through a surface biomimetic approach. Part of this scheme was created with BioRender.com. (b) To mimic the SARS-CoV-2 infection process, the VMP decorated with RBD protrusions were administered to the lung through oropharyngeal aspiration. After internalization via the interaction with ACE2, VMP could trigger robust systemic immune responses, stimulate the production of broad-spectrum neutralizing antibodies, and thereby elicit a broad and cross-reactive immune response against SARS-CoV-2 infection. Furthermore, similar to the SARS-CoV-2 infection process, VMP administration could induce sex-bias differences in ACE2 activity. This scheme was created with BioRender.com. (a) The VMP carrying a substantial number of RBD protrusions were fabricated through a surface biomimetic approach. Part of this scheme was created with BioRender.com. (b) To mimic the SARS-CoV-2 infection process, the VMP decorated with RBD protrusions were administered to the lung through oropharyngeal aspiration. After internalization via the interaction with ACE2, VMP could trigger robust systemic immune responses, stimulate the production of broad-spectrum neutralizing antibodies, and thereby elicit a broad and cross-reactive immune response against SARS-CoV-2 infection. Furthermore, similar to the SARS-CoV-2 infection process, VMP administration could induce sex-bias differences in ACE2 activity. This scheme was created with BioRender.com. Efficient RBD Display on the VMP Surface Programmed by Biotin–Streptavidin Interaction The RBD protein derived from the B.1.617.2 (Delta) variant is capable of inducing high levels of cross-neutralizing antibodies and conferring protection against SARS-CoV-2 lineages. To induce broad-spectrum immunogenicity, B.1.617.2 (Delta) variant RBD protrusions were intensively displayed on the exterior surface of a lipid core, which was composed of 1,2-dioleoyl-3-trimethylammoniumpropane (DOTAP), cholesterol (Chol), dioleoylphosphatidylethanolamine (DOPE), and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[biotinyl­(polyethylene glycol)-2000] (DSPE-PEG2000-biotin). The viral size is closely related to host entrance processes. Only viruses with unique size are able to bind to the host receptors, and thus enter into the host cells. Therefore, solid lipid nanoparticles (SLN) with a size of ∼120 nm, the average diameter of SARS-CoV-2 virions, were fabricated employing a microfluidic technology. Microfluidics can manipulate fluids within channels in dimensions of tens to hundreds of micrometers and, as a result, provide precise control toward particle physicochemical properties. , In the co-flow microfluidic device, the lipid solution flowed into the inner capillary while the aqueous solution of poly­(vinyl alcohol) (PVA, 1%, w/v) was pumped into the space between the inner and outer capillary (Figure a). The lipid molecules were self-assembled into SLN as water diffused into the ethanol phase. During this process, the biotin moieties in DSPE-PEG2000-biotin tend to locate on the SLN exterior surface as a result of the hydrophilicity of PEG and hydrophobicity of DSPE. To enhance the capability of presenting antibodies, a biotin–streptavidin interaction was employed for the particle functionalization. The biotin–streptavidin system is a strong noncovalent interaction with an association constant of 1013–1015 M–1, enabling efficient RBD display on the VMP surface. Since streptavidin is a tetrameric protein with four biotin-binding sites, multiple biotinylated particles might be entrapped by a tetravalent streptavidin, resulting in particle aggregation. Hence, to prevent particle aggregation, streptavidin was first mixed with RBD protrusion (Avi-His-Tag, biotin-labeled, B.1.617.2, delta variant), forming a streptavidin–RBD complex; SLN immobilized with biotin moieties were then bound to the unoccupied binding site of streptavidin in the complex; multiple particles captured by one streptavidin molecule was prevented because of the hindrance resistance provided by the bound RBD protrusion and particle. Efficient RBD display on the VMP surface programmed by biotin–streptavidin interaction. (a) Schematic illustration of the SLN fabrication process with a microfluidic co-flow device and the following immobilization of B.1.617.2 (Delta) variant RBD onto the SLN surface programmed by streptavidin–biotin interaction. (b) The intensity size distribution curves of SLN and VMP. (c–e) The effect of surface immobilization with RBD protrusion on the particle size (c), PDI (d), and zeta potential (e) (Student’s t-test; n = 3; **P < 0.01, n.s., not significant). (f) The transmission electron microscope images of SLN and VMP. (g, h) The measured conjugation efficiency (g) and mass fraction (h) of RBD protrusion in VMP (n = 3). The average diameter of SLN measured by dynamic light scattering was 125.5 nm with a polydispersity index (PDI) of 0.131 ± 0.019, benefiting from the precise control of microfluidics toward the physicochemical properties of particles (Figure b–d). VMP displayed a unimodal size distribution at 135.3 ± 3.5 nm and a comparable PDI (0.165 ± 0.036) with that of SLN, indicating there was bare particle aggregation observed during the conjugation process. After functionalization, the zeta potential decreased (P < 0.01) from +47.4 mV to +33.2 mV (Figure e); the lipid core was surrounded by a gray layer, as observed with a transmission electron microscope; it was easier for the formed VMP to adhere to each other due to the surrounding layer (Figure f). All of these results demonstrated the streptavidin–RBD complex was successfully conjugated to the SLN surface. The conjugation efficiency, the mass ratio of immobilized and originally added Spike S1 RBD, measured with Micro BCA was 52.0% (Figure g). The corresponding mass fraction was 3.87%, which was defined as the amount of Spike S1 RBD immobilized per unit mass of VMP (Figure h). Because of the superior binding affinity of streptavidin to biotin, the mass fraction of bound Spike S1 RBD in our prepared VMP was significantly higher than those reported previously, which might potentially improve the elicited immune response. , Superior Pulmonary Biodistribution and Retention of VMP Mediated by the Functionalization with RBD Protrusion We hypothesized that the fabricated VMP carrying a significant amount of Spike RBD protrusions could enter host lung epithelium through interaction with ACE2 and subsequently trigger a systemic immune response, induce broad-spectrum neutralizing antibody production, and, as a result, realize a robust and broad suppression against SARS-CoV-2 infection. One of the prerequisites for this assumption is that a significant amount of VMP could be effectively delivered to the host lung, which is more focused on the feasibility of the administration method rather than on the interaction between ACE2 and VMP. Therefore, we here sought to evaluate the nanoparticle biodistribution among organs via labeling them with Cyanine5.5 (Cy5.5) in C57BL/6 mice. The mice were administered with phosphate-buffered saline (PBS, control), SLN (10 mg/kg), low-dose VMP (VMP-low, 1 mg/kg), and high-dose VMP (VMP-high, 10 mg/kg), respectively, via oropharyngeal aspiration, followed by execution with terminal euthanasia and organ extraction after 1 day (Figure S1a). Before evaluating their biodistribution, the biocompatibility of nanoparticles was tested via measuring the lung weight and lung/body weight ratio and performing an inflammatory evaluation. No statistically significant changes were observed for lung weight and lung/body weight ratio at 1 day post nanoparticle administration compared with that of the PBS group (Figure S1b,c), showing oropharyngeal aspiration exposure of nanoparticles was not associated with detectable pulmonary edema 24 h following dosing. , Nanoparticle (SLN, VMP-low, and VMP-high) treatment did not elevate inflammatory infiltrate levels across the lung parenchyma compared with the PBS group (Figure S1d,e). These results demonstrate the exceptional biocompatibility of the fabricated nanoparticles in C57BL/6 mice with a dosage as high as 10 mg/kg. The biodistribution of fabricated nanoparticles (SLN and VMP) in C57BL/6 mice was evaluated via fluorescent staining of the extracted organs (Figures and S2). To track the Cy5.5-labeled nanoparticles (red), the cell nuclei and cytomembranes were stained with 4′,6-diamidino-2-phenylindole (DAPI, blue) and rhodamine-phalloidin (green), respectively. For the pulmonary retention analysis, the whole lung was segmented into three areas: trachea, bronchiole, and parenchyma. Bare SLN (10 mg/kg) was retained and distributed throughout the lung at 1 day post administration, while a significant (P < 0.01) amount of VMP in the high-dose group (10 mg/kg) not only deposited in trachea and bronchiole but also diffused into the parenchyma area (Figure a and b). No obvious pulmonary deposition was observed for the VMP-low group, possibly due to the nanoparticle depletion at 1 day post exposure. Since VMP could efficiently suppress SARS-CoV-2 pseudovirion entry into ACE2-expressing cells, its superior biodistribution and retention after inhalation could lead to a potential blockade against SARS-CoV-2 infection in the lung. , As a result, the fabricated VMP are a promising candidate for the next-generation viral inhibitors. Superior pulmonary biodistribution and retention of VMP mediated by functionalization with the RBD protrusion. (a) Representative fluorescent staining images of whole lung, tracheal, bronchiole, and parenchymal sections from C57BL/6 mice at 1 day post pulmonary administration with PBS, SLN (10 mg/kg), VMP-low (1 mg/kg), and VMP-high (10 mg/kg). (b) Quantification of Cy5.5-labeled nanoparticle pixel intensity normalized to nuclei in lung parenchyma from C57BL/6 mice at 1 day post pulmonary administration with PBS (n = 3), SLN (10 mg/kg, n = 4), VMP-low (1 mg/kg, n = 4), and VMP-high (10 mg/kg, n = 4) (one-way ANOVA with posthoc Bonferroni’s test; *P < 0.05, **P < 0.01, n.s., not significant). (c–f) Quantification of Cy5.5-labeled nanoparticle pixel intensity normalized to nuclei in liver (c), kidney (d), spleen (e), and brain (f) sections from C57BL/6 mice at 1 day post pulmonary administration with PBS (n = 3), SLN (10 mg/kg, n = 4), VMP-low (1 mg/kg, n = 4), and VMP-high (10 mg/kg, n = 4) (one-way ANOVA with posthoc Bonferroni’s test; *P < 0.05, n.s., not significant). Off-target biodistribution of the nanoparticles (SLN and VMP) was tracked in the other main organs, including the liver, kidney, spleen, and brain (Figures c–f and S2). SLN accumulation was not observed in these organs at 1 day post administration, attributable to its rapid clearance. For the VMP-high group (10 mg/kg), obvious (P < 0.05) nanoparticle distribution was observed in the liver, an organ which accumulated plenty of mononuclear phagocytes (Figures c and S2). The administered VMP might be translocated from the lung to the bloodstream and subsequently delivered to the liver via circulation. By contrast, negligible VMP was observed in the kidney, spleen, and brain at 1 day post administration (Figures d–f and S2). These results were in accordance with the evidence that phagocytosis via mononuclear phagocytes and hepatobiliary elimination are two main clearance pathways for foreign materials. VMP Administration Elicited Robust and Broad Systemic Immune Response with Sex-Bias Differences in Pulmonary ACE2 Activity in K18-hACE2 Mice As the prepared VMP were proven to be effectively delivered to and retained in the host lung for at least 1 day, we further evaluated the immunogenicity of VMP by measuring the elicited immune response. The VMP-elicited immune response was measured in K18 human angiotensin-converting enzyme 2 transgenic (K18-hACE2) mice, an animal model that is highly susceptible to SARS-CoV-2 infection and thus suitable for vaccine identification and characterization. To this end, four groups of K18-hACE2 mice were dosed with PBS (control), SLN (10 mg/kg), VMP-low (1 mg/kg), and VMP-high (10 mg/kg), respectively, via oropharyngeal aspiration, which was followed by scheduled euthanasia and organ, blood, and bronchoalveolar lavage fluid (BALF) harvest after 1, 7, and 14 days (Figure S3a). Similarly, the biocompatibility of nanoparticles (SLN and VMP) in K18-hACE2 mice was evaluated before measuring the elicited immune response. The K18-hACE2 mice that were dosed with nanoparticles (SLN and VMP) gained weight gradually and revealed a similar body weight profile with the PBS group within 2 weeks, showing no gross health impairment by nanoparticle exposure (Figure S3b). Compared with the PBS group, the pulmonary administration with nanoparticles (SLN and VMP) did not recruit inflammatory macrophages, lymphocytes, neutrophils, eosinophils, and basophils, and thus, the level of total white blood cells remained at the basal level within 2 weeks post administration (Figure S3c–h). Furthermore, the nanoparticle (SLN and VMP) groups revealed comparable tumor necrosis factor alpha (TNF-α) level with that of the PBS group (Figure S3i); elevated interleukin 6 (IL-6) was observed for the SLN group at 1 day post administration (Figure S3j), while VMP inhalation did not lead to obvious increment, indicating the surface functionalization with RBD protrusion reduced the nanoparticle potential of inducing cytokine storm. For histological analysis, H&E-stained lung sections from VMP groups demonstrated normal architecture and bare infiltrates of inflammatory cells compared with the PBS group (Figures S3k and S4). The Immunoglobulin G (IgG) responses in K18-hACE2 mice sera were measured to evaluate the capacity of fabricated VMP for humoral immunity inducement. VMP-high elicited broad systemic immune response in females at 14 days post pulmonary administration: exponential (P < 0.001) increase of Spike-specific IgG antibodies was observed against SARS-CoV-2 B.1.617.2 (Delta) variant, in accordance with the derivation of Spike RBD protrusions carried on the VMP surface (Figure a); broad systemic immune response was induced against antigens from wild-type SARS-CoV-2 and its variants of concern, including D614G, B.1.1.7 (Alpha), B.1.351 (Beta), P.1 (Gamma), and XBB.1 (Omicron) (Figures b–g and S5). The administered VMP carried Spike RBD of 6.75 μg for females (data not shown) in the VMP-high group, which was calculated based on its mass fraction and average mice body weight. As no obvious IgG response was reported to be induced by the immunization with Spike S1 RBD of 20 μg, VMP-high elicited not only broad but also robust systemic immune response at 14 days post pulmonary administration. Furthermore, the antibodies induced by VMP reveal a potential to target antigens from Omicron subvariants, consistent with previous findings that Delta variant RBD nanoparticles led to cross-neutralizing antibodies against BA.1, BA.2, BA.3, BA.2.12.1, BA.2.75, and BA.4/5 variants in guinea pigs. Comparatively, there was no immune response observed for SLN and VMP-low groups in K18-hACE2 mice at 14 days after administration. VMP administration elicited negligible IgG increment against SARS-CoV-2 nucleocapsid since it carried Spike RBD protrusions instead of nucleocapsid on its surface (Figure h). Moreover, bare Immunoglobulin M (IgM) and Immunoglobulin A (IgA) increment were measured in serum or BALF (Figure S6). VMP administration elicited robust and broad systemic immune response in K18-hACE2 mice. (a) The intensity of binding IgG antibody induced by PBS (n = 4, male and female), SLN (10 mg/kg, n = 6, male and female), VMP-low (1 mg/kg, n = 6, male and female), and VMP-high (10 mg/kg, n = 3, female) against SARS-CoV-2 Spike from B.1.617.2 (Delta) subvariant (one-way ANOVA with posthoc Bonferroni’s test; ***P < 0.001, n.s., not significant). (b and c) The intensity of binding IgG antibody induced by PBS (n = 4, male and female), SLN (10 mg/kg, n = 6, male and female), VMP-low (1 mg/kg n = 6, male and female), and VMP-high (10 mg/kg, n = 3, female) against SARS-CoV-2 RBD (b) and Spike (c) from wild-type strain (one-way ANOVA with posthoc Bonferroni’s test; ***P < 0.001, n.s., not significant). (d) The intensity of binding IgG antibody induced by PBS (n = 4, male and female), SLN (10 mg/kg, n = 6, male and female), VMP-low (1 mg/kg, n = 6, male and female), and VMP-high (10 mg/kg, n = 3, female) against SARS-CoV-2 Spike from D614G subvariant (one-way ANOVA with posthoc Bonferroni’s test; ***P < 0.001, n.s., not significant). (e and f) The intensity of binding IgG antibody induced by PBS (n = 4, male and female), SLN (10 mg/kg, n = 6, male and female), VMP-low (1 mg/kg, n = 6, male and female), and VMP-high (10 mg/kg, n = 3, female) against B.1.1.7 (Alpha, e) and B.1.351 (Beta, f) variant SARS-CoV-2 (one-way ANOVA with posthoc Bonferroni’s test; ***P < 0.001, n.s., not significant). (g) The intensity of binding IgG antibody induced by PBS (n = 4, male and female), SLN (10 mg/kg, n = 6, male and female), VMP-low (1 mg/kg, n = 6, male and female), and VMP-high (10 mg/kg, n = 3, female) against SARS-CoV-2 Spike from P.1 (Gamma) subvariant (one-way ANOVA with posthoc Bonferroni’s test; ***P < 0.001, n.s., not significant). (h) The intensity of binding IgG antibody induced by PBS (n = 4, male and female), SLN (10 mg/kg, n = 6, male and female), VMP-low (1 mg/kg, n = 6, male and female), and VMP-high (10 mg/kg, n = 3, female) against SARS-CoV-2 nucleocapsid from wild-type strain (one-way ANOVA with posthoc Bonferroni’s test; n.s., not significant). The pulmonary administration of VMP-high in male K18-hACE2 mice was accompanied by rapid onset of clinical signs, such as closed eyes, decreased activity, weakness, and incoordination, while those clinical signs were not observed with female mice from the VMP-high group. Four male K18-hACE2 mice dosed with VMP-high were euthanized at ∼2 h post pulmonary administration. Those mice revealed a comparable terminal lung/body weight ratio to the control group (Figure S7a), male K18-hACE2 mice without nanoparticle administration, indicating negligible pulmonary edema was induced by pulmonary administration of VMP-high, at least before euthanasia. Furthermore, VMP-high treatment did not elevate inflammatory infiltrate levels across the lung parenchyma in male K18-hACE2 mice compared with the control group (Figure S7b and c). To further indicate the potential reasons for sex-bias differences in VMP-induced morbidity in K18-hACE2 mice, immunofluorescence staining was performed to visualize the human ACE2 (green) expression pattern in the extracted lung sections (Figure a and b). DAPI (blue) was used as a counterstain to indicate the cell nuclei, while RBD protrusion displayed on the VMP surface was stained with Alexa Fluor 647 (red). A comparable expression level of human ACE2 was observed in male and female K18-hACE2 mice at baseline, which remained constant in females after VMP-high administration (Figure a and c). However, VMP-high induced a dramatically (P < 0.05) decreased ACE2 expression at ∼2 h post pulmonary administration in males compared with the baseline (Figure b and d). ACE2 restoration predominantly depends on transcriptional and translational upregulation, which is a slow process, typically requiring several days rather than hours. Furthermore, a similar ACE2 decrease is reported in men after SARS-CoV-2 infection, which is supposed to contribute to higher mortality in men versus women. ACE2 was reported to catalyze the conversion of angiotensin II (Ang II) to angiotensin 1–7 (Ang1–7), providing protective effects against acute lung failure during Coronavirus disease 2019 (COVID-19) infection. Besides, the reduced endogenous ACE2 expression is associated with a dramatic decrease in estrogen receptor α (Erα) expression, likewise leading to loss of organ protection in males. Thus, the decreased ACE2 expression in males might be the reason for sex-bias differences in VMP-induced morbidity. Sex-bias differences in pulmonary ACE2 activity in K18-hACE2 mice. (a) Representative immunofluorescence staining images of lung sections from K18-hACE2 mice at 1 day post pulmonary administration with PBS, SLN (10 mg/kg), VMP-low (1 mg/kg), and VMP-high (10 mg/kg). (b) Representative immunofluorescence staining images of lung sections from male K18-hACE2 mice without administration and at ∼2 h post pulmonary administration with VMP-high (10 mg/kg). (c) Quantification of ACE2 pixel intensity normalized to nuclei in lung sections from K18-hACE2 mice at 1 day post pulmonary administration with PBS (n = 2), SLN (10 mg/kg, n = 3), VMP-low (1 mg/kg, n = 3), and VMP-high (10 mg/kg, n = 3) (one-way ANOVA with posthoc Bonferroni’s test; n.s., not significant). (d) Quantification of ACE2 pixel intensity normalized to nuclei in lung sections from male K18-hACE2 mice without administration (n = 3) and at ∼2 h post pulmonary administration with VMP-high (10 mg/kg, n = 4) (one-way ANOVA with posthoc Bonferroni’s test; *P < 0.05). (e) Quantification of RBD pixel intensity normalized to nuclei in lung sections from K18-hACE2 mice at 1 day post pulmonary administration with PBS (n = 2), SLN (10 mg/kg, n = 3), VMP-low (1 mg/kg, n = 3), and VMP-high (10 mg/kg, n = 3) (one-way ANOVA with posthoc Bonferroni’s test; n.s., not significant). (f) Quantification of RBD pixel intensity normalized to nuclei in lung sections from male K18-hACE2 mice without administration (n = 3) and at ∼2 h post pulmonary administration with VMP-high (n = 4, 10 mg/kg) (one-way ANOVA with posthoc Bonferroni’s test; *P < 0.05). Unlike in C57BL/6 mice, bare VMP was retained and distributed throughout the lung at 1 day post administration of VMP-high in female K18-hACE2 mice (Figure e), which can be ascribed to the higher ACE2 expression levels in K18-hACE2 mice, and the resultant rapid cell internalization and following metabolism of VMP. Compared with the control group, mice without administration, a significant (P < 0.05) amount of VMP in the high-dose group was deposited in the male parenchyma area at ∼2 h post pulmonary administration (Figures f and S8), which indicated the successful pulmonary delivery of nanoparticles through oropharyngeal aspiration. In summary, we fabricated VMP with substantial RBD protrusions on the particle surface to elicit a broad and cross-reactive immune response against SARS-CoV-2 infection. The lipid core of VMP was prepared by a microfluidic platform, which sufficiently controlled the nanoprecipitation process and fabricated SLN with a unimodal and narrow size distribution. Enabled by the high affinity of the biotin–streptavidin interaction, VMP were decorated with abundant RBD protrusions on each particle surface, replicating the Delta variant of SARS-CoV-2. After oropharyngeal aspiration exposure, the fabricated VMP revealed superior pulmonary biodistribution and retention for at least 1 day in C57BL/6 female mice, which was primarily mediated by the functionalization with RBD protrusion. Due to multiple conformation-dependent epitopes and highly conserved regions in the RBD, inhalation of high-dose VMP induced both robust neutralizing IgG antibodies against SARS-CoV-2 B.1.617.2 (Delta) variant and a broad systemic immune response against wild-type SARS-CoV-2 and its variants of concern, including D614G, B.1.1.7 (Alpha), B.1.351 (Beta), P.1 (Gamma), and XBB.1 (Omicron). The VMP represent a highly promising candidate for the development of next-generation viral inhibitors and immune stimulating vaccines targeting the rapid variant mutations of SARS-CoV-2. In addition, our studies show that the female K18-hACE2 mice were less vulnerable to the adverse effects induced by the inhaled VMP compared with male mice. We speculate that this can be ascribed to the pulmonary ACE2 activity differences: inhalation of VMP-high dramatically decreased ACE2 expression in males, whereas that in females was comparable to the baseline. Since these phenotypic differences in ACE2 between biological sexes may also contribute to worse outcomes and higher mortality rates in males than in females, our study also sheds light on the sex-bias differences in severity after COVID-19 infection, warranting further comprehensive research. Furthermore, additional inhalation studies are required to address the potential adverse effects caused by the oropharyngeal aspiration exposure of VMP in transgenic mice. Supplementary Material